The Reconstruction Of Laser-based Real-time Machine Vision Measuring Data

Measurement technology is developing in the pace with technology improvement and mechanical manufacturing development. Especially with the rapid progress of electronics technology, measurement technology tends to high precision and efficient. Thesis uses a technique of laser-based vision to measure a subject's profile and gives some steps to prove precision. The technique uses a mapping function theory. Firstly, CCDs grasp the line image projected by the laser and get the 3-D information of an object, secondly, we switch the CCDs'image data to space point data based on mapping function theory, then, the measuring is completed. During the realization, the important work and innovations are:a) Thesis uses minimum error analysis at establishing mapping function and achieving the subpixel measuring, improves the precision of measuring greatly.b) Thesis discusses histograms, threshold division, stripe thinning at image processing, and establishes a filter method to the noise during merging the left and right CCDs'info.c) Thesis notices the calibration inside the equipments and the whole system to improve precision and gives a new correct method of the laser detecter sensor's level and vertical degree.d) Thesis uses the progress and the notices presented by software engineering at software programming, thus ensure the software's quick, low error, and high code's reusing identity. This thesis also includes the C++ Builder language's gigantic power at exploitation. It also gives new measuring software that adopting both top-down and bottom-up methods.